Yann LeCun says world models, not LLMs, will drive the next AI breakthrough
Overview
AMI Labs founder Yann LeCun says the next AI breakthrough will come from systems that understand and predict the physical world, not from large language models.
Speaking in Seoul, he said current AI cannot understand reality, citing robots that cannot complete tasks without training and the lack of L5 self-driving cars.
LeCun argues that world models can learn physical laws and causal relationships, letting them predict the results of actions. This is his claim, reported by IThome; no supporting results were given in the report.
Written by AI from the articles below · updated Oct 12, 7:37 AM ET
Check the sources:
Article timeline
The articles in this story. Times are ET.
- IThome · AINewsYann LeCun says the next AI breakthrough will come from world models, not LLMs
AIAMI Labs founder Yann LeCun said in Seoul on Monday that the next AI breakthrough will come from systems that understand and predict the physical world rather than from large language models. He said today's AI lacks the ability to understand reality, citing robots that cannot complete tasks without training and the absence of L5 self-driving cars, and argued that world models can learn physical laws and causal relationships to predict the results of actions.
Heat trend
Not enough continuous observations to show a trend yet.